Feature · Call analytics
Call analytics for volume, outcomes and cost per call.
PyAI logs call length, who answered, cost and failed requests without reading the conversation.
What PyAI gives you for call analytics
- PyAI call analytics records duration, language, outcome and spend for every finished voice agent call, split across three Console views: Reliability, Usage and cost by meter, and filterable Request logs.
- AMD history counts human answers, machine pickups and unknown outcomes with their decision time at $0.004 a call, charting daily volume across a 7-, 30- or 90-day window you choose.
- An extraction schema turns each call's result, such as a booked appointment date, into JSON sent as a call-results event, so reports count outcomes without parsing transcripts.
How does one call turn into numbers on the reports?
Choose an example and run it. The lines are an example script, not a recording, showing what PyAI writes when each call ends.
Pick a moment, then press play
How does call analytics collect data from each call?
Call analytics usually splits into numbers about calls and analysis of conversations. PyAI covers the numbers side from one record per call.
Call ends
An Omni agent finishes an inbound or outbound call, or JustCall pushes a completed call to the private webhook URL.
Record committed
PyAI saves duration, language, metadata and transcript, deduplicates by call id, and keeps a stereo recording only when the agent allows recording.
Outcome attached
AMD adds who answered with confidence and decision time, and an extraction schema adds the business result as JSON fields.
Counters move
Usage and cost, Reliability and Request logs pick up the call, while webhooks and app destinations receive the events.
API request errors
The failed row appears in Request logs with its status code, ready for an engineer to copy the request id into a ticket.
Voicemail picks up
AMD history stores a machine outcome with its evidence, lowering that day's human answer share in the outcome mix.
Caller drops after 20 seconds
PyAI still writes the record with the real 20-second duration, so short calls count toward volume and spend rather than disappearing.
Who reads call analytics data, and for what decision?
Five teams use these call numbers. Each team opens its own tab. All five count from the same record per call, so finance and operations always see the same totals.
Reaching people, not voicemail
AMD history separates human answers from machines and unknowns day by day, so a campaign lead can move dialing to the days with a higher human answer share.
Cost per call
A four-minute Omni call costs $0.24 in Omni and telephony minutes, plus $0.004 if AMD ran, and Usage and cost shows the monthly total per meter.
Failures by status
Filtering Request logs to error statuses for last Tuesday lists each broken request with an id that support can trace.
Outcomes by campaign
An extraction schema writes booked or not booked into JSON, and a HubSpot destination files it against the contact.
Outbound volume records
Call records with durations and AMD outcomes give a dated trail of who was dialed and who picked up, which teams keep for their own reviews.
Call analytics questions people ask
What metrics does call analytics software track?
Call reporting software usually tracks volume, duration, peak hours, answer outcomes, abandonment and cost per call. PyAI records duration and language per call, human versus machine outcomes and daily volume in AMD history, spend by meter in Usage and cost, and failed requests by status in Request logs.
Does PyAI include call tracking for ads?
No. PyAI does not assign tracking numbers or attribute calls to ad campaigns, which is the job of marketing call tracking tools. PyAI measures the call itself instead: duration, answering machine decisions, extracted outcomes, spend per meter and request reliability for Omni voice agents and imported JustCall calls.
How much does call reporting cost in PyAI?
PyAI lists no separate charge for the Analytics tab, Request logs or call records, which come with every project. AMD costs $0.004 per call when used. Omni voice agent calls cost $0.05 per minute plus $0.01 per minute for telephony, with no platform fee and $50 of starter credit.
How do I see my full phone call history?
On a phone or in Microsoft Teams, call history lives in that app, which PyAI does not cover. For PyAI voice agents, AMD history lists every detection decision for 7, 30 or 90 days with cursor paging, and each Omni call record is fetchable by id with duration, transcript and recording.
Related features
PyAI is voice infrastructure. It hands you the audio and transcript, and does not replace your own models or review tools.